Summary Loop10

🚀 Summary Loop Model

A text summarization model based on GPT2 architecture, capable of generating summaries for given documents.

🚀 Quick Start

In the right panel, you can try out the model (although it only handles a short sequence length). Just enter the document you want to summarize in the panel on the right.

📦 Installation

The model (based on a GPT2 base architecture) can be loaded in the following way:

from transformers import GPT2LMHeadModel, GPT2TokenizerFast

model = GPT2LMHeadModel.from_pretrained("philippelaban/summary_loop10")
tokenizer = GPT2TokenizerFast.from_pretrained("philippelaban/summary_loop10")

💻 Usage Examples

Basic Usage

document = "Bouncing Boulders Point to Quakes on Mars. A preponderance of boulder tracks on the red planet may be evidence of recent seismic activity. If a rock falls on Mars, and no one is there to see it, does it leave a trace? Yes, and it's a beautiful herringbone-like pattern, new research reveals. Scientists have now spotted thousands of tracks on the red planet created by tumbling boulders. Delicate chevron-shaped piles of Martian dust and sand frame the tracks, the team showed, and most fade over the course of a few years. Rockfalls have been spotted elsewhere in the solar system, including on the moon and even a comet. But a big open question is the timing of these processes on other worlds — are they ongoing or did they predominantly occur in the past?"

tokenized_document = tokenizer([document], max_length=300, truncation=True, return_tensors="pt")["input_ids"].cuda()
input_shape = tokenized_document.shape
outputs = model.generate(tokenized_document, do_sample=False, max_length=500, num_beams=4, num_return_sequences=4, no_repeat_ngram_size=6, return_dict_in_generate=True, output_scores=True)
candidate_sequences = outputs.sequences[:, input_shape[1]:] # Remove the encoded text, keep only the summary
candidate_scores = outputs.sequences_scores.tolist()

for candidate_tokens, score in zip(candidate_sequences, candidate_scores):
    summary = tokenizer.decode(candidate_tokens)
    print("[Score: %.3f] %s" % (score, summary[:summary.index("END")]))

Example Output

[Score: -0.084]  Here's what you need to know about rockfalls
[Score: -0.087]  Here's what you need to know about these tracks
[Score: -0.091]  Here's what we know so far about these tracks
[Score: -0.101]  Here's what you need to know about rockfall

📚 Documentation

You can access more information, access to the scoring function, the training script, or an example training log on the Github repo: https://github.com/CannyLab/summary_loop

📄 License

This project is licensed under the Apache-2.0 license.